4D-ARE MCP Server

From the README

# 4D-ARE: Attribution-Driven Agent Requirements Engineering

> Build LLM agents that explain *why*, not just *what*.

## The Problem

Your LLM agent has full data access and executes flawlessly. But when asked:

> "Why is our customer retention rate only 56%?"

It returns a list of metrics instead of a causal explanation:

```
Retention rate: 56%
Visit frequency: 2.1
Cross-sell rate: 28%
...
```

**This is the Attribution Gap** - agents can report *what* happened, but struggle to explain *why*.

## The Solution

4D-ARE provides a framework for building agents that trace causal chains through 4 dimensions:

```
Results (What happened)
    ↓
Process (What we did)
    ↓
Support (What resources we had)
    ↓
Long-term (What environment we're in)
```

Instead of a metric dump, you get:

```
Results: Retention dropped to 56% (target: 80%)
    ↑ caused by
Process: Visit frequency declined 23%, cross-sell rate low
    ↑ constrained by
Support: Staffing ratio at 68%, understaffed
    ↑ driven b